HAART is often complicated by difficulties with adherence and drug toxicity. ACTG 384 is designed to examine several strategies for maintaining long-term HIV suppression while minimizing the risk of non-adherence and drug toxicity. The strategic questions addressed in ACTG 384 are questions being asked by many physicians currently treating HIV-infected subjects. These questions are:(1) Is it better to start therapy with two NRTIs and a protease inhibitor or two NRTIs and an NNRTI?; (2) Is a four-drug regimen consisting of two NRTIs, a protease inhibitor, and an NNRTI superior to two consecutive three-drug regimens, each consisting of two different NRTIs and either a protease inhibitor or an NNRTI?; (3) Is there a preferred order in which dual NRTI regimens should be administered in subjects who require a change in anti-HIV therapy? Question 1: There has been no direct comparison between two NRTIs and a protease inhibitor vs. two NRTIs and an NNRTI. Several studies have shown the antiviral benefits of combining two NRTIs with a protease inhibitor. However, these therapies have often been associated with toxicities, including metabolic abnormalities (lipodystrophy, diabetes) and sometimes the emergence of broad cross-resistance to other protease inhibitors. Whether to use protease inhibitors in initial regimens or to hold them in reserve for later therapy is one of the fundamental questions confronting physicians and HIV-infected individuals today. In ACTG 384, NFV + two NRTIs will be compared to EFV + two NRTIs. The combination of ZDV, ddI, and the NNRTI nevirapine reduced plasma HIV-1 RNA >2.0 log in previously untreated individuals, and 50% of individuals receiving the 3-drug combination had plasma HIV-1 RNA levels < 20 copies/mL after 1 year. EFV is an experimental NNRTI that appears more potent than nevirapine in vitro, and in a recent phase II study (DMP 266-005), the combination of EFV + ZDV + 3TC suppressed plasma HIV-1 RNA to <400 copies/mL for at least 6 months in nearly all subjects. Question 2: A four-drug regimen consisting of two NRTIs + NFV + EFV will be compared to two consecutive three-drug regimens consisting of two NRTIs + either NFV or EFV followed by two different NRTIs + NFV for those who initially receive EFV or by two different NRTIs + EFV for those who initially receive NFV. The four-drug regimen may have greater antiviral potency, but at the greater risk of toxicity. Moreover, the four-drug regimen may provide fewer options for salvage therapy in subjects who develop virologic failure. The consecutive three-drug regimen strategy may have less toxicity and may provide a potent follow-up regimen for individuals who experience drug toxicity or virologic failure during the first regimen. Question 3: ACTG 384 will assess which treatment strategy provides the best long-term approach to safe and effective HIV management. Among the six treatment arms, individuals randomized to Arms A,B, and E will receive ddI 400 mg qd + d4T 40 mg q12h (with appropriate dose reductions for persons weighing <60 kg), and individuals randomized to Arms C, D, and F will receive ZDV 300 mg q12h + 3TC 150 mg q12h in the form of the ZDV/3TC fixed-dose combination. Individuals in Arms A, B, and E who must switch regimens because of virologic failure or drug intolerance will change their nucleoside regimen to ZDV + 3TC; individuals in Arms C, D, and F who must switch regimens because of virologic failure or drug intolerance will change their nucleoside regimen to d4T + ddI. It has been suggested that the use of certain NRTIs (e.g. ZDV) may affect the subsequent phosphorylation of other NRTIs. Thus, the order to NRTI use may be important in selecting the best strategy for initial management. ACTG 384 will not only provide the best long-term approach to safe and effective HIV management, but will also compare which strategy is optimal initially.